add model

.gitignore

@@ -0,0 +1 @@
+checkpoint-*/

README.md

@@ -0,0 +1,73 @@
+---
+language:
+- en
+license: apache-2.0
+tags:
+- generated_from_trainer
+datasets:
+- ag_news
+metrics:
+- accuracy
+model_index:
+- name: distilbert-base-uncased-agnews
+  results:
+  - dataset:
+      name: ag_news
+      type: ag_news
+      args: default
+    metric:
+      name: Accuracy
+      type: accuracy
+      value: 0.9473684210526315
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# distilbert-base-uncased-agnews
+
+This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the ag_news dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.1652
+- Accuracy: 0.9474
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 3e-05
+- train_batch_size: 32
+- eval_batch_size: 8
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_steps: 1000
+- num_epochs: 2.0
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Accuracy |
+|:-------------:|:-----:|:----:|:---------------:|:--------:|
+| 0.1916        | 1.0   | 3375 | 0.1741          | 0.9412   |
+| 0.123         | 2.0   | 6750 | 0.1631          | 0.9483   |
+
+
+### Framework versions
+
+- Transformers 4.8.2
+- Pytorch 1.8.1+cu111
+- Datasets 1.8.0
+- Tokenizers 0.10.3

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.0,
+    "eval_accuracy": 0.9473684210526315,
+    "eval_loss": 0.16520710289478302,
+    "eval_runtime": 10.786,
+    "eval_samples_per_second": 704.618,
+    "eval_steps_per_second": 88.077
+}

config.json

@@ -0,0 +1,36 @@
+{
+  "_name_or_path": "distilbert-base-uncased",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForSequenceClassification"
+  ],
+  "attention_dropout": 0.1,
+  "dim": 768,
+  "dropout": 0.1,
+  "hidden_dim": 3072,
+  "id2label": {
+    "0": "World",
+    "1": "Sports",
+    "2": "Business",
+    "3": "Sci/Tech"
+  },
+  "initializer_range": 0.02,
+  "label2id": {
+    "Business": 2,
+    "Sci/Tech": 3,
+    "Sports": 1,
+    "World": 0
+  },
+  "max_position_embeddings": 512,
+  "model_type": "distilbert",
+  "n_heads": 12,
+  "n_layers": 6,
+  "pad_token_id": 0,
+  "problem_type": "single_label_classification",
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "transformers_version": "4.8.2",
+  "vocab_size": 30522
+}

log.log

@@ -0,0 +1,18 @@
+Training dataset length:
+108000
+Validation dataset length:
+12000
+Test dataset length:
+7600
+Current performance:
+Eval:
+{'eval_loss': 1.3898906707763672, 'eval_accuracy': 0.21558333333333332, 'eval_runtime': 17.4618, 'eval_samples_per_second': 687.213, 'eval_steps_per_second': 85.902}
+Test:
+{'eval_loss': 1.3894953727722168, 'eval_accuracy': 0.21947368421052632, 'eval_runtime': 10.7033, 'eval_samples_per_second': 710.062, 'eval_steps_per_second': 88.758}
+Best trial:
+BestRun(run_id='0', objective=0.9483333333333334, hyperparameters={'learning_rate': 3e-05, 'num_train_epochs': 2, 'per_device_train_batch_size': 32, 'warmup_steps': 1000})
+Training complete performance:
+Eval:
+{'eval_loss': 0.16314448416233063, 'eval_accuracy': 0.9483333333333334, 'eval_runtime': 17.4366, 'eval_samples_per_second': 688.209, 'eval_steps_per_second': 86.026, 'epoch': 2.0}
+Test:
+{'eval_loss': 0.16520710289478302, 'eval_accuracy': 0.9473684210526315, 'eval_runtime': 10.786, 'eval_samples_per_second': 704.618, 'eval_steps_per_second': 88.077, 'epoch': 2.0}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40a8f0e24751f591a63839b7f462428c62c58378130bb47db3dddfbe0c96abf0
+size 267869335

special_tokens_map.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}

tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "distilbert-base-uncased", "tokenizer_class": "DistilBertTokenizer"}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.0,
+    "eval_accuracy": 0.9473684210526315,
+    "eval_loss": 0.16520710289478302,
+    "eval_runtime": 10.786,
+    "eval_samples_per_second": 704.618,
+    "eval_steps_per_second": 88.077
+}

trainer_state.json

@@ -0,0 +1,144 @@
+{
+  "best_metric": 0.16314448416233063,
+  "best_model_checkpoint": "../results/distilbert-base-uncased-agnews/run-31/checkpoint-6750",
+  "epoch": 2.0,
+  "global_step": 6750,
+  "is_hyper_param_search": true,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.5e-05,
+      "loss": 0.6831,
+      "step": 500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3e-05,
+      "loss": 0.27,
+      "step": 1000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.7391304347826085e-05,
+      "loss": 0.2275,
+      "step": 1500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.4782608695652175e-05,
+      "loss": 0.2233,
+      "step": 2000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.217391304347826e-05,
+      "loss": 0.2019,
+      "step": 2500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 0.1916,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9411666666666667,
+      "eval_loss": 0.17406368255615234,
+      "eval_runtime": 17.0218,
+      "eval_samples_per_second": 704.977,
+      "eval_steps_per_second": 88.122,
+      "step": 3375
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.6956521739130433e-05,
+      "loss": 0.1747,
+      "step": 3500
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.4347826086956522e-05,
+      "loss": 0.1382,
+      "step": 4000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.173913043478261e-05,
+      "loss": 0.1389,
+      "step": 4500
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.130434782608697e-06,
+      "loss": 0.1269,
+      "step": 5000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 0.1299,
+      "step": 5500
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.913043478260869e-06,
+      "loss": 0.1328,
+      "step": 6000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.3043478260869566e-06,
+      "loss": 0.123,
+      "step": 6500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9483333333333334,
+      "eval_loss": 0.16314448416233063,
+      "eval_runtime": 16.9602,
+      "eval_samples_per_second": 707.537,
+      "eval_steps_per_second": 88.442,
+      "step": 6750
+    },
+    {
+      "epoch": 2.0,
+      "step": 6750,
+      "total_flos": 1.0849763309273088e+16,
+      "train_loss": 0.20914978592484085,
+      "train_runtime": 832.6813,
+      "train_samples_per_second": 259.403,
+      "train_steps_per_second": 8.106
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9483333333333334,
+      "eval_loss": 0.16314448416233063,
+      "eval_runtime": 17.4366,
+      "eval_samples_per_second": 688.209,
+      "eval_steps_per_second": 86.026,
+      "step": 6750
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9473684210526315,
+      "eval_loss": 0.16520710289478302,
+      "eval_runtime": 10.786,
+      "eval_samples_per_second": 704.618,
+      "eval_steps_per_second": 88.077,
+      "step": 6750
+    }
+  ],
+  "max_steps": 6750,
+  "num_train_epochs": 2,
+  "total_flos": 1.0849763309273088e+16,
+  "trial_name": null,
+  "trial_params": {
+    "learning_rate": 3e-05,
+    "num_train_epochs": 2,
+    "per_device_train_batch_size": 32,
+    "warmup_steps": 1000
+  }
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c27892004ed1c4922eb92e0f06efecef465902cd27c4bccaa474412b62545b03
+size 2799